Wikipedia:Wikipedia Signpost/Single/2024-03-02

For prior Signpost coverage, see see Section 230 report (February 2023)

On the Wikimedia-l mailing list, two members of the Wikimedia Foundation's "Global Advocacy" team drew attention to

important hearings happening this week at the United States Supreme Court.

The hearings on two cases that will be crucial for Wikimedia have just started: NetChoice, LLC v. Paxton and Moody v. NetChoice, LLC. Both cases are challenges to state laws in Texas and Florida, which impact content moderation on social media websites. [...] As they are written, these laws prohibit website operators from banning users or removing speech and would generally risk Wikipedia’s volunteer-led systems of content moderation. That’s because these laws were designed to prevent social media platforms from engaging in politically motivated content moderation, but were drafted so broadly that they would also impact Wikipedia. The case is also important beyond the impact it might have on our projects. It represents a scenario that is part of a trend globally, where governments introduce legislation to address harms from big tech actors, yet Wikimedia ends up as the dolphin inadvertently caught in the net."

The Foundation has previously weighed in on these cases with an amicus brief and several blog posts, and is present at the current hearings "in person talking to stakeholders and observing the proceedings. We expect the Court to rule this year and will be providing updates as we know more."

Asked about the worst-case scenario (from a Wikimedia perspective), Stan Adams of the Global Advocacy team elaborated:

"Perhaps the worst long-term outcome would be if several other states or even the US Congress replicated the Texas or Florida laws. If those laws were enforced against Wikipedia editors or the Foundation – say, for editors' regular work of removing content that is inaccurate, unsourced, or that violates NPOV policies – it could become increasingly difficult to operate and maintain Wikipedia."

However,

"based on what I observed at the Court yesterday [February 26, mentioning comments by justice Brett Kavanaugh in particular], I think most of the Justices would be reluctant to uphold the Texas and Florida laws. That said, these cases won't be the end of legislative attempts to regulate social media and other venues for expression online – I expect to see the Court considering more cases like these as states continue to enact laws that raise First Amendment questions in the online context."

– H

The Wikimedia Foundation advised on Meta-Wiki that –

A vote to ratify the charter for the Universal Code of Conduct Coordinating Committee (U4C) was held from 19 January until 2 February 2024 via SecurePoll. Voting is now closed. Thank you to all who voted. The result was 1249 voters in support and 420 voters opposed. 69 voters did not choose an option. Voter statistics and a summary of voter comments will be published soon.

You can find more information on the U4C's purpose and scope here. – AK

Are more changes afoot for the Requests for adminship process? Open proposals from Phase I include the following (some others were already rendered unsuccessful).

• Proposal 2: Add a reminder of civility norms at RfA
• Proposal 3: Add three days of discussion before voting (trial)
• Proposal 3b: Make the first two days discussion-only (trial)
• Proposal 4: Prohibit threaded discussion (trial)
• Proposal 5: Add option for header to support limited-time adminship (trial)
• Proposal 6: Provisional adminship via sortition
• Proposal 6b: Trial adminship
• Proposal 7: Threaded General Comments
• Proposal 8: Straight vote (trial)
• Proposal 10: Unbundling 90% of blocks
• Proposal 12: Abolish the discretionary zone and crat chats
• Proposal 12b: Abolish crat chats and allow discretionary relisting
• Proposal 13: Admin elections
• Proposal 14: Suffrage requirements
• Proposal 16: Allow the community to initiate recall RfAs
• Proposal 16b: Require a reconfirmation RfA after X years
• Proposal 16c: Community recall process based on dewiki
• Proposal 16d: Community recall process initiated by consensus
• Proposal 17: Have named Admins/crats to monitor infractions
• Proposal 18: Normalize the RfB consensus requirements
• Proposal 20: Make RFA an internal non public process
• Proposal 21: Reduce threshold of consensus at RfA
• Proposal 22: Change the name from RFA to "Nominations For Adminship"

Phase I is still open, and you may weigh in with your thoughts here: Wikipedia:Requests for adminship/2024 review. – B

From February 19 to February 23, 2024, "a group of 21 Wikimedians, academics, and practitioners" met at the Rockefeller Foundation's Bellagio Center in Northern Italy "to draft an initial research agenda on the implications of artificial intelligence (AI) for the knowledge commons." The aim is "to focus attention (and therefore resources) on the vital questions volunteer contributors have raised, including the promise, as well as risks and negative impacts, of AI systems on the open Internet." The agenda is available on Meta-Wiki, together with a brief report on the meeting.

Members of the "Wikimedia AI" Telegram group expressed their surprise about hearing about this effort first from organizations outside the Wikimedia movement, and about the fact that the term "open source" isn't mentioned in the document (despite open-source AI being an important topic of debate in AI currently, and WMF's general commitments to the use of open source software). While the announcement appears to be speaking on behalf of "volunteer contributors", the "Wikimedians" involved in drafting the document appears to have consisted exclusively of Wikimedia Foundation staff (largely from its Research department), according to the attendee list. Wikimedia Foundation CEO Maryana Iskander subsequently clarified that this "effort to contribute to a shared research agenda on AI [...] was created by a small group working in the open who rushed to publish a ‘bad first draft’ that will benefit from more input."

In other AI-related news, the Wikimedia Foundation recently received a $2.2 million grant from the Sloan Foundation (a longtime supporter) for the purpose of "leverag[ing] AI for the benefit of Wikipedia's readers and contributors, including tools to address vandalism" over the next three years. (These funds come on top of a $950,000 grant announced in April 2023 by WMF's own Wikimedia Endowment for "building and strengthening AI and machine learning infrastructure on Wikipedia and Wikimedia projects", similarly highlighting "the development of algorithms to measure the quality of Wikipedia articles and machine learning models that help catch incidents of vandalism on Wikimedia projects.")

– AK H

• Annual reports:
  • Wikimedians for Sustainable Development,
  • Wikipedia & Education User Group,
  • Wikimedians of Republic of Srpska, Wikimedia Belgium,
  • Wikisource Community User Group,
  • Wikimedia South Africa,
  • Wikimedia Community Ireland User Group,
  • Wiki Advocates Philippines User Group.
  • Les sans pagEs

• New administrator: The fourth user to receive administrator rights this year is Sdkb, whose request for adminship was closed as successful with 265 votes in support and 2 against. The Signpost welcomes the English Wikipedia's newest administrator.
• Articles for Improvement: This week's Article for Improvement is Economy of South America, followed by Community service (beginning March 4). Please be bold in helping improve these articles!
• New sources at the Wikipedia Library: The WMF and Taylor & Francis have announced that users of The Wikipedia Library will have free access to all T&F and Routledge journals going forward. If you've never used the Wikipedia Library, it's an incredible resource for research! The Signpost covered the library in detail last September.

A monthly overview of recent academic research about Wikipedia and other Wikimedia projects, also published as the Wikimedia Research Newsletter.

Reviewed by Bri and Tilman Bayer

A Nature paper titled "Online Images Amplify Gender Bias"^[1] studies:

"gender associations of 3,495 social categories (such as 'nurse' or 'banker') in more than one million images from Google, [English] Wikipedia and Internet Movie Database (IMDb), and in billions of words from these platforms"

As summarized by Neuroscience News:

This pioneering study indicates that online images not only display a stronger bias towards men but also leave a more lasting psychological impact compared to text, with effects still notable after three days.

This was a two-part research paper in which the authors:

• examined text and images from the Internet for gender bias
• examined the responses of experimental subjects who were exposed to text and images from the Internet

While the paper's main analyses focus on Google, the authors replicated their findings with text and image data from Wikipedia and IMDb.

For the first part, images were retrieved from Google Search results for 3,495 social categories drawn from WordNet, a canonical database of categories in the English language. These categories include occupations—such as doctor, lawyer and carpenter—and generic social roles, such as neighbour, friend and colleague. Faces extracted from these images (using the OpenCV library) were tagged with gender by workers recruited via Amazon Mechanical Turk. The reliability of tagging was validated against the self-identified gender from a "canonical set" of celebrity portraits culled from IMDb and Wikipedia.^{[supp 1]}

For the replication analysis with English Wikipedia (relegated mainly to the paper's supplement), an analogous set of images was derived using another existing Wikipedia image dataset,^{[supp 2]} whose text descriptions yielded matches for 1,523 of the 3,495 WordNet-derived social categories (For example, we retrieve the Wikipedia article with the title ‘Physician’ for the social category physician: https://en.wikipedia.orgview_html.php?sq=Envato&lang=en&q=Physician).

• All images with faces in the current version of the article physician
• 
• 
• 

To measure gender bias in a corpus of text from e.g. Google News, the authors use word embeddings (a computational natural language processing technique) trained on that corpus. Specifically, their method (adapted from a 2019 paper) assigns a number to each category (e.g. doctor, lawyer or carpenter) that captures the extent to which [the word for this category] co-occurs with textual references to either women or men [in the corpus]. This method allows us to position each category along a −1 (female) to 1 (male) axis, such that categories closer to −1 are more commonly associated with women and those closer to 1 are more commonly associated with men [in the corpus]. [...] The category ‘aunt’, for instance, falls close to −1 along this scale, whereas the category ‘uncle’ falls close to 1 along this scale. The authors interpret any deviation of this "gender association" value from 0 as evidence of "gender bias" for a particular category. Figure 1 in the paper illustrates this in case of Google News for a list of occupations. There, the three categories with the largest male bias appear to be "football player", "philosopher", and "mechanic", and the three categories with the largest female bias "cosmetologist", "ballet dancer", and "hairstylist". In the figure, the category closest to being unbiased (0) in the Google News text was "programmer". Overall though, texts from Google News exhibit [only] a relatively weak bias towards male representation, with an average score of 0.03.

In case of Wikipedia text, this gender association of a particular WordNet category was determined using a pre-trained word embedding model of Wikipedia available in Python’s gensim package, which was built using the GloVe method to analyze a 2014 corpus of 5.6 billion words from Wikipedia. Somewhat concerningly, this description by the authors is inconsistent with the gensim documentation, which states that this 5.6 billion token corpus was not based on Wikipedia alone, but on "Wikipedia 2014 + Gigaword". According to the original GloVe paper,^{[supp 3]} "Gigaword 5 [...] has 4.3 billion tokens", meaning that it would form a much bigger part of that corpus than Wikipedia. (The GloVe authors also observed that Wikipedia's entries are updated to assimilate new knowledge, whereas Gigaword is a fixed news repository with outdated and possibly incorrect information; the corpus contains newswire text dating back to 1994.)

In other words, the Nature study's conclusions about Wikipedia text might not be valid. Assuming they are though, they might seem vaguely reassuring for Wikipedians (and perhaps somewhat in contrast with earlier research about textual gender bias on Wikipedia): Using several different variants of the model (with different word embedding dimensions), respectively, 57% (50D), 59% (100D), 57.6% (200D), and 54% (300D) of categories [are] male-skewed, with an average strength of gender association below 0.06 (recall that the authors describe the corresponding value of 0.03 for Google News as a relatively weak bias). The story is different for images, though:

images over Wikipedia are significantly skewed toward male representation. 80% of categories are male-skewed according to images over Wikipedia (p < 0.0001, proportion test, n = 495, two-tailed). [...] Including all 1,244 categories in our analysis continues to show a strong bias toward male representation in Wikipedia images (with 68% of faces being male, p < 0.00001). [...] Wikipedia content can appear to be neutral in its gender associations if one focuses only on text, whereas examining Wikipedia images from the same articles can reveal a different reality, with evidence of a strong bias toward male representation and a stronger bias toward more salient gender associations in general.

For the second part (which did not involve Wikipedia directly), the researchers

... conducted a nationally representative, preregistered experiment that shows that googling for images rather than textual descriptions of occupations amplifies gender bias in participants’ beliefs.

To measure participants' gender bias after they had completed the googling task, an implicit association test (IAT) methodology was used, which supposedly reveals unconscious bias in a timed sorting task. In the researchers' words, "the participant will be fast at sorting in a manner that is consistent with one's latent associations, which is expected to lead to greater cognitive fluency [lower measured sorting times] in one's intuitive reactions." Specifically, the IAT variant used was designed to detect the implicit bias towards associating women with liberal arts and men with science. The test measured how long participants took to associate a particular word or image (e.g. "Girl", "Engineering", "Grandpa", "Fashion") with either the male/female or science/liberal arts categories.

The labeling of text descriptions was performed by other humans recruited via Amazon Mechanical Turk. Both the test subject, and the labelers, were adults from the United States, and the test subjects were screened to be representative of the U.S. population to include a nearly 50/50 male/female split (none self identified as other than those two categories). The experiment focused on a sample of 22 occupations, e.g. immunologist, harpist, hygienist, and intelligence analyst.

Some test subjects were given a task related to occupation-related text prior to the IAT, and some were given a task related to images. The task was either to use Google search to retrieve images of representative individuals in the occupation, or Google search to retrieve a textual description of the occupation. A control group performed an unrelated Google search. Before the IAT was performed, the test subjects were required to indicate on a sliding scale, for each of the occupations, "which gender do you most expect to belong to this category?" The test was performed again a few days later with the same test subjects.

On the second test, subjects exposed to images in the first test had a stronger IAT score for bias than those exposed to text.

The experimental part of the study depends partly on IAT and partly on self-assessment to detect priming, and there are concerns about replicability concerning the priming effect, and the validity and reliability of IAT. Some of the concerns are described at Implicit-association test § Criticism and controversy. It seemed that the authors recognized this (We acknowledge important continuing debate about the reliability of the IAT), and in their own study found that the distribution of participants' implicit bias scores [arrived at with IAT] was less stable across our preregistered studies than the distribution of participants' explicit bias scores, and discounted the implicit bias scores somewhat.

The conclusion drawn by the researchers, based partly but not entirely on the different IAT scores of experimental subjects, was that of the paper title: "images amplify gender bias" — both explicitly as determined by the subject's assignments of occupation to gender on a sliding scale, and implicitly as determined by reaction times measured in the IAT.

The paper opens with the (rather thinly referenced) observation that "Each year, people spend less time reading and more time viewing images". Combined with the finding that searching for occupation images on Google amplified participants' gender biases, this forms an "alarming" trend according to the study's lead author (Douglas Guilbeault of UC Berkeley's Haas School of Business), as quoted by AFP on "the potential consequences this can have on reinforcing stereotypes that are harmful, mostly to women, but also to men".

The researchers also determined, apart from experimental subjects, that the Internet – represented singularly by Google News – exhibits a strong gender bias. It was unclear to this reviewer how much of the reported Internet bias is really "Google selection bias". Based on these findings, the authors go on to speculate that "gender biases in multimodal AI may stem in part from the fact that they are trained on public images from platforms such as Google and Wikipedia, which are rife with gender bias according to our measures".

• See the page of the monthly Wikimedia Research Showcase for videos and slides of past presentations.
• Submissions are open until April 22, 2024 for Wiki Workshop 2024, to take place on June 20, 2024. The virtual event will be the eleventh in this annual series (formerly part of The Web Conference), and is organized by the Wikimedia Foundation's research team with other collaborators. The call for contributions asks for 2-page extended abstracts which will be "non-archival, meaning we welcome ongoing, completed, and already published work."

Other recent publications that could not be covered in time for this issue include the items listed below. Contributions, whether reviewing or summarizing newly published research, are always welcome.

Compiled by Tilman Bayer

From the abstract:^[2]

From the abstract: "[...] measuring stereotypes is difficult, particularly in a cross-cultural context. Word embeddings are a recent useful tool in natural language processing permitting to measure the collective gender stereotypes embedded in a society. [...] We considered stereotypes associating men with career and women with family as well as those associating men with math or science and women with arts or liberal arts. Relying on two different sources (Wikipedia and Common Crawl), we found that these gender stereotypes are all significantly more pronounced in the text corpora of more economically developed and more individualistic countries. [...] our analysis sheds light on the “gender equality paradox,” i.e. on the fact that gender imbalances in a large number of domains are paradoxically stronger in more developed/gender equal/individualistic countries."

To determined "the relative contribution of residents from each country to each language [version of Wikipedia]", the author (a researcher at CNRS) used the Wikimedia Foundation's "WiViVi" dataset which provides the percentage of pageviews per country for a given language Wikipedia. This data is somewhat outdated (last updated in 2018) and also, for the goal of measuring contribution (rather than consumption), the separate Geoeditors dataset might have been worth considering (which provides the number of editors per country, although with - somewhat controversial - privacy redactions).

From the abstract:^[3]

"for many people around the world, [Wikipedia] serves as an essential news source for major events such as elections or disasters. Although Wikipedia covers many such events, some events are underrepresented and lack attention, despite their newsworthiness predicted from news value theory. In this paper, we analyze 17 490 event articles in four Wikipedia language editions and examine how the economic status and geographic region of the event location affects the attention [page views] and coverage [edits] it receives. We find that major Wikipedia language editions have a skewed focus, with more attention given to events in the world’s more economically developed countries and less attention to events in less affluent regions. However, other factors, such as the number of deaths in a disaster, are also associated with the attention an event receives."

Relatedly, a 2016 paper titled "Dynamics and biases of online attention: the case of aircraft crashes"^[4] had found:

that the attention given by Wikipedia editors to pre-Wikipedia aircraft incidents and accidents depends on the region of the airline for both English and Spanish editions. North American airline companies receive more prompt coverage in English Wikipedia. We also observe that the attention given by Wikipedia visitors is influenced by the airline region but only for events with a high number of deaths. Finally we show that the rate and time span of the decay of attention is independent of the number of deaths and a fast decay within about a week seems to be universal.

From the abstract:^[5]

"[...] we present FAMuS, a new corpus of Wikipedia passages that report on some event, paired with underlying, genre-diverse (non-Wikipedia) source articles for the same event. Events and (cross-sentence) arguments in both report and source are annotated against FrameNet, providing broad coverage of different event types. We present results on two key event understanding tasks enabled by FAMuS: source validation -- determining whether a document is a valid source for a target report event -- and cross-document argument extraction -- full-document argument extraction for a target event from both its report and the correct source article. "

From the abstract of this preprint by a group of authors from Google Research and Georgia Institute of Technology:^[6]

"... we formally present the task of Open-domain Visual Entity recognitioN (OVEN), where a model need to link an image onto a Wikipedia entity with respect to a text query. We construct OVEN-Wiki by re-purposing 14 existing datasets with all labels grounded onto one single label space: Wikipedia entities. OVEN challenges models to select among six million possible Wikipedia entities, making it a general visual recognition benchmark with the largest number of labels. Our study on state-of-the-art pre-trained models reveals large headroom in generalizing to the massive-scale label space. We show that a PaLI-based auto-regressive visual recognition model performs surprisingly well, even on Wikipedia entities that have never been seen during fine-tuning."

From the paper:^[7]

"... through a case study of comparing human [Wikidata] items of two countries, Vietnam and Germany, we propose several reasons that might lead to the existing biases in the knowledge contribution process. [...]
We chose Germany and Vietnam as subjects based on three primary considerations. Firstly, both nations have comparable population sizes. Secondly, the editors who speak the predominant languages of each country maintain their distinct Wiki communities on Wikidata. [...]
The first analysis we did was comparing different components of Wikidata pages between pages in two countries. The components we are comparing are labels, descriptions, claims, and sitelinks. For a single Wikidata page, label is the name that this item is known by, while description is a short sentence or phrase that also serves disambiguate purpose. [...] In the dataset we collected, there are 290,750 people who have citizenship of Germany, and there are only 4,744 people who have citizenship of Vietnam. [...] German pages on average had 13 more labels, 5 more descriptions and 7 more claims compared to Vietnamese pages. While surprisingly, Vietnamese pages had slightly more sitelinks, the difference according to effect size was negligible.
The second analysis focused on the edit history of Wikidata items. [...] we quantified the attention metric into five features: Number of total edits, number of human edits, number of bot edits, and number of distinct bot and human edits. [...] in all the five features the [difference in means between the German and Vietnamese Wikidata human pages] is significant and in terms of bot activity and total activity, the effect size is beyond medium threshold (0.5).

From the abstract:^[8]

... an extended case study is developed on the (re)construction of a major pollution event (the [1952] Great Smog of London). Critical discourse analysis of intertextuality (connections between texts through hyperlinking and other shared patterning) is utilised to move from a focus on micro level practices to macro and meta level findings on the ordering of Wikipedia and its interactions with other institutions. Findings evidence a layered, self-referencing formation across texts, favouring the interests of established institutions and providing limited opportunity for marginalised groups to interact with sustained (re)constructions of the Great Smog. Comparison to a previous study of the constructed memory of a crisis (the London Bombings 2005) reveals dynamics across Wikipedia that lead to an emphasis on connecting (re)constructions to institutional traditions rather than the potential usefulness of such (re)construction for those at higher risk of negative outcomes arising from repeated crises.

Supplementary references and notes:

As first reported by The Scottish Sun, and then shared by The National, a few computers from the Scottish Parliament's headquarters might have been used to edit the Wikipedia articles of several MSPs from all over the political spectrum. Both newspapers correctly remind that all kinds of edits made to a specific page are automatically archived in its history: in this case, the original inquiry found that the IP addresses of some unregistered users apparently trace back to Holyrood. If confirmed, this would represent an evident breach not only of the parliamentary code of conduct, but also Wikipedia's rules on conflict of interest.

The main subject involved in the inquiry is Scottish Liberal Democrats leader, Alex Cole-Hamilton. Back in February 2021, the MSP received widespread backlash after being caught insulting his fellow politician Maree Todd, who was then serving as the Minister for Children and Young People, during an online institutional meeting. On February 16 of the same year, user Alex B4 added a short mention of the incident, which included a National article that further criticized Cole-Hamilton for his inappropriate apologies to Todd. However, on February 24, an IP user stepped in to remove references to the incident entirely: then, they proceeded to cut down part of the information provided on Cole-Hamilton's expenses for his campaign ahead of the 2016 elections, calling them "ad hominem attacks". All of the content was eventually restored by Alex himself on March 7. Since then, another IP user deleted some of the references about the 2021 incident again, replacing them with an unsourced statement claiming that Cole-Hamilton wrote a letter saying sorry to Todd, before apologizing to her in person – a statement that is still up at the time of writing this piece.

Other notable MSPs involved in the report are incumbent Scottish First Minister and SNP leader, Humza Yousaf, who was added to a list of "notable alumni" of the US-backed International Visitor Leadership Program; Reform UK Scotland leader, Michelle Ballantyne, who had an entire section about her political controversies removed from her article; and, finally, incumbent Minister for Cabinet and Parliamentary Business, George Adam, whose love for football club St Mirren F.C. was further highlighted on his page – curiously, Adam was the only one who immediately responded upon being contacted by The Sun, and it looks like he just had a good laugh out of the "incident". – O, B and RTH

In a recent episode of CBS-hosted comedy panel show After Midnight, aired on February 12, 2024, host Taylor Tomlinson arranged a special wikirace as part of one of the show's mini-games, Wikipedia Link. In the occasion, her fellow comedians Vinny Thomas, Riki Lindhome and Rob Huebel took turns to guess how many clicks it takes to go from Snoop Dogg to the Great Depression on the English Wikipedia.

Tomlinson introduced the game by deeming Wikipedia as "humanity's CliffsNotes", as the three panelists then shared increasingly unorthodox guesses, ranging from weed to The Grapes of Wrath. Although it was Huebel who eventually found the right number of articles needed to complete the race, specifically five (including the two aforementioned pages), Tomlinson revealed quite an unexpected pattern: from Snoop, to Peanuts, to Howdy Doody, to Wonder Bread, to the 1939 New York World's Fair, to the Great Depression. However, in the comments below the video extract available on the show's YouTube channel, several users have stated there are even shorter paths connecting the two pages.

No matter who is right, it's safe to say Taylor and the rest of the After Midnight staff deserve a shout-out for helping popularize the wikirace trend and, by extension, Wikipedia as a whole. – O

The MIT Technology Review has published a profile of the Wikimedia Foundation's CTO, Selena Deckelmann. The main focus of the piece is how Deckelmann sees the place of Wikipedia in the age of chatbots:

Deckelmann argues that Wikipedia will become an even more valuable resource as nuanced, human perspectives become harder to find online. But fulfilling that promise requires continued focus on preserving and protecting Wikipedia’s beating heart: the Wikipedians who volunteer their time and care to keep the information up to date through old-fashioned talking and tinkering. Deckelmann and her team are dedicated to an AI strategy that prioritizes building tools for contributors, editors, and moderators to make their work faster and easier, while running off-platform AI experiments with ongoing feedback from the community. “My role is to focus attention on sustainability and people,” says Deckelmann. “How are we really making life better for them as we’re playing around with some cool technology?”

However –

Today Deckelmann sees a newer sustainability problem in AI development: the predominant method for training models is to pull content from sites like Wikipedia, often generated by open-source creators without compensation or even, sometimes, awareness of how their work will be used. “If people stop being motivated to [contribute content online],” she warns, “either because they think that these models are not giving anything back or because they’re creating a lot of value for a very small number of people—then that’s not sustainable.” At Wikipedia, Deckelmann’s internal AI strategy revolves around supporting contributors with the technology rather than short-circuiting them. The machine-learning and product teams are working on launching new features that, for example, automate summaries of verbose debates on a wiki’s "Talk" pages (where back-and-forth discussions can go back as far as 20 years) or suggest related links when editors are updating pages. “We’re looking at new ways that we can save volunteers lots of time by summarizing text, detecting vandalism, or responding to different kinds of threats,” she says.

The article also discusses the potential need for Wikipedia to meet its readers elsewhere online, naming the Foundation's Wikipedia ChatGPT plugin as an example. – AK

• Wikidata and Wikipedia as arbiters of truth: The Australian edition of The Spectator, noting that Wikipedia feeds internet search results on Google, Yahoo!, or Bing, as well as question-answering systems such as Siri, Amazon Alexa, or AI language models, wonders "how safe it is to crowdsource the truth".
• You're fired: Controversial radio talk show Coast to Coast AM recently claimed that "a cabal of hardcore skeptics and UFO debunkers has surreptitiously seized control of hundreds of Wikipedia pages". What's more, a change.org petition has been set up to "Demand Wikipedia to Dismiss Editors Suppressing UAP Phenomenon Information": one of its over 600 supporters stated that the platform "should show all sides of an issue without demanding so-called 'scientific evidence'".
• Where's my cut?: According to infamous news website BNN Breaking, under the surface of Wikipedia lies the "Billion-Dollar Web of Deception, Lies & Elite Manipulation", as evidenced by "deliberate distortion" in the articles about BNN Breaking's founder Gurbaksh Chahal, his wife, and BNN Breaking itself.
• Endorsed by law professor: Pepperdine Law associate professor Jacob Charles actually trusts Wikipedia, so much that he said you should all go read the article about State constitutions in the United States right away.
• Endorsed by librarians: UC Berkeley Libraries appears to have coined the term "wikiphiliacs" in an announcement of a Valentine's Day special edit-a-thon. The always-reliable Google Trends tells us that it's likely to be truly novel.
• But also not endorsed by librarians (do your own research): The Georgia State University Library recently hosted a seminar on how to use books, journals and portals like Gale eBooks for research, "instead of relying on Wikipedia".
• Endorsed by meme as "generally reliable source": A Wikipedia-themed meme went viral on Twitter/X last week, reaching over 13 million views. Based on the "midwit" meme format featuring three Wojak characters of low, average and high intelligence, the meme mocks the supposedly naive concern "Nooooo anyone can edit it!" (expressed by the "midwit" character) with the supposedly more valid take that "Wikipedia is generally a reliable source" (expressed by both the low and high intelligence characters).
• Wikipedia: The Telos of the internet: In an article for the The Daily Northwestern, Scott Hwang praised Wikipedia's ad-free, community-centered model and strict rules on neutrality and verifiability, deeming the platform as the real-life version of the The Hitchhiker's Guide to the Galaxy.
• Argentine edit-a-thon: Wikimedian and UNLP professor Fernando Archuby was recently interviewed on Argentine radio station Radio Antena Libre, as he hosted a work-shop in his local community to help people learn how to edit Wikipedia and improve the articles related to the natural and cultural heritage of northern Patagonia.

• Wish I'd thought of that, whatever it is: The Guardian recently reported about the overnight success of Harlesden-based developer TJ Gardner, who last year made roughly £‎280,000 out of creating apps that are basically made of a start button, a click counter, "some lo-fi acoustic beats looping endlessly in the background" and... a cute creature from Wikimedia Commons.
• Online Images Amplify Gender Bias: The Independent has reported on a recent study which proved how online pictures can inadvertently amplify gender bias – see related Signpost coverage in Recent research.
• Robot accountants versus Wikipedia: David Pierce reported for The Verge on how robots.txt works and is especially robust on several platforms, including Wikipedia, and how the rise of AIs could be a problem for "robot accounting".
• Российский encyclopedia considers copying from Wikipedia: Meduza, Russian Wikinews and other sources reported on some statements from Russian presidential aide Vladimir Medinsky, who suggested copying Wikipedia in its entirety, as well as Russian fork Ruviki, to expand another analog of the platform, Znanie. Wiki.
• Yoruba Wikipedia hits 25 million views in 2023: Nigerian newspaper The Nation has reported that the Yoruba Wikipedia hit 25 million annual views in 2023, up from about 17 million annual views in 2021, thus officially becoming the most-visited Yoruba-language website.
• Florida man nominated for deletion: The discussion that led to the (temporary) deletion of Miami-Dade County mayoral election candidate Manny Cid seemingly sparked backlash from the Republican politician himself, according to the local press.
• The Pacific Northwest has its own encyclopedia: The Seattle Times has shed a light on Historylink, an online encyclopedia of Washington state history that has CC-BY content, but can't exactly be deemed as a "fork"... since it's actually older than Wikipedia itself.

• They have the time, the life experience, and they know what they want to write about: AFP let us discover the stories of the Czech seniors who have been involved in "Seniors Write Wikipedia", a project started in 2013 by the national branch of the WMF to help elderly people learn how to edit the Czech Wikipedia. One of the participants, retired librarian Jirina Kadnerova, summed up her enthusiasm about contributing to the platform by saying, "When I write about something, I want to lay my hands on it".
• Wikipedia use rejected in U.S. court: Law360 has noted that a judge on the U.S. Court of International Trade decided to reject attempts from the Department of Commerce to use Wikipedia articles in support of their plans to expand vertical shaft engine tariffs. And if you're wondering about that vertical engine thing which is currently a redlink, we think they are gasoline motors for small machines, of the type made by Zongshen in China and Briggs and Stratton in the United States.

• Wikipedia's AI worries: The Financial Times covered Wikipedia co-founder Jimmy Wales' latest interview for Japan's The Nikkei. In the interview, Wales said a well-known AI self-censored after beginning to reply about the Tiananmen Square massacre.
• This should not be so routine: In a formal complaint to Columbia University, the head of diversity, equity and inclusion at the university's Irving Medical Center, Alade McKen, was alleged to have plagiarized approximately 20% of his doctoral thesis from Wikipedia, among other sources. The news were reported by Fox News and The Washington Free Beacon – with the latter even showing the matching passages, including "a near-verbatim facsimile of Wikipedia's entry on Afrocentric education" running about two pages in the dissertation.

We are deeply saddened to inform you that American Wikipedia volunteer Vami suddenly passed away on February 13, 2024, aged 24, just one day after making his last edits on the platform. He was a contributor on Commons, Wikidata, Meta, the English Wikipedia, and elsewhere.

Fellow Wikipedians sawyer-mcdonell and AirshipJungleman29 wrote a brief note about Vami's accomplishments:

Vami_IV joined Wikipedia in November 2014 and soon became a prolific editor across a wide variety of topics, including buildings, wildfires, video games, and the history of Europe, Texas, and Latin America. In total, he brought ten articles to featured status, including two worthy of the Four Award, and had nominated two more at WP:FAC at the time of his passing; he also brought numerous articles to A or GA-class, including a year-long project on Simón Bolívar. Vami often collaborated in these efforts with other editors; he was active at both WikiProject Military history, at which he was a coordinator between 2021 and 2023, and WikiProject Women in Red, for which he created 167 articles on primarily Latin American women. He also contributed extensively to WP:CCI and was active at the Wikimedia Discord, where he was unfailingly witty and generous with his time, especially with newer editors.

He will be remembered for his editing talents, his personality, and his dedication to the project, exemplified by his philosophy of Completionism:

"As a Wikipedia editor, I understand that my goal is the destruction of my purpose."

On behalf of the whole Signpost staff, we would like to send our condolences to Vami's family and loved ones. You can leave respectful comments of remembrance on his talk page.

• These lists exclude the Wikipedia main page, non-article pages (such as redlinks), and anomalous entries (such as DDoS attacks or likely automated views). Since mobile view data became available to the Report in October 2014, we exclude articles that have almost no mobile views (5–6% or less) or almost all mobile views (94–95% or more) because they are very likely to be automated views based on our experience and research of the issue. Please feel free to discuss any removal on the Top 25 Report talk page if you wish.

For the January 25 – February 25 period, per this database report.

The first round of the 2024 WikiCup, English Wikipedia's annual editing competition, has just concluded, and it's turning out to be an exciting one! This year saw 135 participants, the highest since 2017. Due to the high number of participants, a total of 30 points were needed to qualify to the next round, the most since 2014, and the third-highest of all time.^[1] Due to a tie for the last spot, 67 users advanced (normally 64), including yours truly, the defending champ, who made it as the 30th spot!^[2]

See also: The WikiCup 2023.

The WikiCup began humbly in 2007 with 12 competitors, scoring being based primarily on edit counts and unique page edits. The competition adopted its present form in 2009, with points awarded for featured articles (FA), lists (FL) and pictures (FP), along with good articles (GA), In the news (ITN) and Did you know (DYK). There were 60 contestants that year and 120 the next, and the contest has taken place annually ever since.

135 users entered – 83 scored points – only 67 advanced. Here we have some notable scores and participants:

• 1st – Generalissima (submissions) – a rookie WikiCup participant who led the pack through two months of competition, scoring 915 points, which included a featured article on John Littlejohn (preacher), a leading 15 good articles and 16 DYKs.
• 2nd – Vami_IV (submissions) – who sadly passed away earlier this month. Scored two FAs (leading the round) along with a number of reviews (tying for the lead in featured content reviews), two GAs and a DYK for 790 points.
• 3rd – AirshipJungleman29 (submissions) – a finalist from last year, scoring 580 points in several categories including an FA on Hö'elün.
• 4th – Sammi Brie (submissions) – another WikiCup veteran and finalist from last year, totaling 420 points, mainly from GAs and DYKs relating to radio and television stations.
• 5th – MaranoFan (submissions) – scoring 351 points, mainly from music articles including an FA on Holidays (Meghan Trainor song); had the largest – and only – featured topic (nine articles) of the round.
• 6th – Skyshifter (submissions) – scored 345 points, mainly from an FA on the game OneShot.
• 7th – voorts (submissions) – totaled 270 points, primarily from an FA on the song Running Out of Time.
• 8th – SounderBruce (submissions) – who's off to a flying start in round 2 with a featured article already, but who placed 8th in round 1 with 247 points, mostly from FLs, GAs and DYKs relating to sports.
• 9th – Hey man im josh (submissions) – active in the featured list process, scoring 235 points while being the leader in both FLs (4) and joint-leader in featured content reviews (with Vami IV).
• 10th – Elli (submissions) – 235 points, mostly from a featured article on the 1964 Illinois House of Representatives election (an interesting election in which over 500 million votes were cast...in a state of 12 million?).

• 11th – Muboshgu (submissions) – a WikiCup veteran who led the first round in ITN entries, scoring 195 points from 15 ITNs and two DYKs.
• 22nd – Lee Vilenski (submissions) – 2020 and 2022's champion; can Lee Vilenski be the first three-time winner in WikiCup history?
• 30th – BeanieFan11 (submissions) – yikes! I'm tryin' to defend my title and this is the best I can do? I'm going to have to step it up if I want to have any chance of being the third back-to-back champion!
• 31st – Ealdgyth (submissions) – the current leader in good article reviews, scoring 65 points on 13 of them.
• 52nd – TonyTheTiger (submissions) – as far as I can tell, the earliest WikiCup participant in this year's competition; was the runner-up in 2010. Can Tony finally do it after 14 years?

Thank you to all those who have been greatly improving Wikipedia in the competition! Good luck to the remaining 67 in your (and my) quest to be the 2024 WikiCup champion in the remaining four rounds until October 31!